一种基于粒子群的多站测向交叉定位改进算法

An Improved Algorithm for Multi-station DF Cross Location Based on Particle Swarm

针对现有的多站测向交叉定位算法定位精度低、迭代初值不易选取等问题,提出了一种基于改进粒子群的多站交叉定位算法。新算法利用多站测向信息构建最小二乘(Least Squares,LS)误差模型,以此作为适应度函数。通过LS定位方法求出目标辐射源位置的粗略解,结合该解限制粒子搜索空间;对粒子群优化(Particle Swarm Optimization,PSO)算法中学习因子参数进行非线性调整,平衡粒子在局部与全局二者之间的寻优能力;在粒子搜索后期结合模拟退火算法中的Metropolis准则,避免粒子无法获得最优定位解。仿真结果表明,新算法不需要设置迭代初值,能有效地对目标进行定位,尤其在测角误差较大时同其他多站测向交叉定位算法相比而言具有更高的定位精度。

To solve the problems of low positioning accuracy and difficult selection of initial iterative values in the existing multi-station direction finding cross-location algorithms,an improved multi-station cross-location algorithm based on particle swarm is proposed.The new algorithm uses the multi-station DF information to build the Least Squares(LS)error model,which is used as the fitness function of the particle swarm optimization algorithm.Firstly,the algorithm combines the principle of LS positioning to obtain a rough solution of the target radiation source position,and the particle search space is limited by the solution;Secondly,the parameters of learning factor in Partical Swarm Optimization(PSO)are adjusted nonlinearly to balance the optimization ability of particles between the local and the global;Finally,in the later stage of iteration,it combines the Metropolis criterion in the simulated annealing algorithm to avoid that particles cannot obtain the optimal positioning solution.The simulation results show that the improved algorithm does not need to set the initial value of iteration,and can effectively locate the target,especially when the measurement error is large,it has higher positioning accuracy than other multi-station DF cross-location algorithms.